Title of Invention

METHOD AND DEVICE FOR ENCODING MPEG-4 VIDEO DATA

Abstract The invention relates to a method and a device for encoding video data consisting of one or several bitstreams according to the MPEG-4 standard. In order to avoid any problem when mapping the video elements into the system elements, when said video bitstreams are encoded using the syntax mode corresponding to the fragmentation of the Video Object Planes (VOPs) contained in said video data into Video Packets (VPs) and of Video Packets into Data Partitions (DPs), a Video Data Partition is mapped into one or more SL packets (SLl, SL2, SL3, SL4...), the first Video Data Partition (DP1) start is always mapped to an SL packet start, and the last SL packet transporting the first Data Partition includes the separation marker and up to 7 subsequent bits of the second Data Partition (DP2) in order to obtain byte alignment, the next SL packet starting on the next bit of the second Data Partition.
Full Text

The present invention relates to a method for encoding video data consisting
of one or several bitstreams according to the MPEG-4 standard. It also relates to a A , i corresponding encoding device and to a coded MPEG-4 signal consisting ofat least a video , ,
■ Uu-'-
bitstream obtained at the output of such an encoding device.
,.. -'^
i
The transmission of audio-visual data on lossy networks, such as the Internet or UMTS radio channels, requires the use of coding techniques that are both efficient in their use of bits and robust against transmission errors. The MPEG-4 standard, which has been designed in this context and provides standardized ways to represent audio-visual objects (AVOs) of natural or synthetic origin, exploits both the temporal and spatial redundancies found in natural and synthetic video sequences. To that purpose, for the three types of Video, Object Planes (VOPs) present in the video stream (intra VOPs, or I VOPs ; predicted VOPs, or P VOPs; bidirectional VOPs, or B VOPs), specific coding techniques are used. These coding techniques, which reduce the bandwidth requirements by removing the redundancy in the video signal, become, when said signal is partitioned into packets in the transport layer, quite sensitive to bit errors and transport-layer packet losses (for example, a single bit error may make the decoding of information impossible, or, owing to a transport-layer packet loss, the predictively-coded motion information in the next packets may become undesirable). Moreover, due to the predictive nature of the encoder, an error which occurs in an I or P VOP tends to propagate to the following P and surrounding B VOPs.
Error resilience is then one of the numerous options offered by the MPEG-4 video standard in order to solve the previously indicated drawbacks : it provides a set of tools which allows to hierarchically sort out the encoded data according to their sensitivilji (this error resilient mode includes inter ail a functionality improving the ability of the decoder to localize an error : it is based on data partitioning, achieved by separating the motion and macro block header information and the texture information, which allows, if the texture information is lost, to use the motion information to conceal such an error by means of a motion compensation of the previously decoded VOP). In order to take advantage of this

feature, the transport layer must take into account the hierarchy information provided by the video layer.
The MPEG-4 video bitstreams are classically composed of a number of elements such as : Video Objects (VOs), Video Object Layers (VOLs), Video Object Planes, Group of VOPs (GOVs), Video Packets (VPs), Video Data Partitions (VDPs), etc..., and MPEG-4 system manipulates entities such as : DecoderSpecificInfo, Access Units (AUs), SL packets. For this, to work properly, how to map the video elemi’s into the system elements is of key importance. In particular, the mapping of Video Data Partitions to system is described here. Video Data Partitions correspond to fragments of Video Packets, in a specific video bitstream syntax mode that enables them for error resilience purposes; specifically, there are two video Data Partitions for each Video Packet. A drawback of the Data Partition syntax is however that it is not byte aligned, i.e. the boundary between the first and second Data Partitions of a Video Packet is not byte aligned: it does not start on a bit multiple of 8. This situation is sub-optimal for an efficient machine implementation and may lead to problems when considering network transport, since network protocols transport bytes (i.e. slices of eight bits).
It is therefore a first object of the invention to propose a method avoiding this drawback.
To this end, the invention relates to a method for encoding video data consisting of one or several bitstreams according to the MPEG-4 standard, in which a specific alignment/fragmentation mechanism is chosen, according to which, when said video bitstreams are encoded using the syntax mode corresponding to the fragmentation of the Video Object Planes (VOPs) contained in said video data into Video Packets (VPs) and of Video Packets into Data Partitions (DPs), a Video Data Partition is moped into one or more SL packets, the first Video Data Partition start is always mapped to an SL packet start even if a large Video Data Partition is spited across several SL packets, and the last SL packet transporting the first Data Partition includes the separation marker (DC marker or Motion Marker depending on VOP type) and up to 7 subsequent bits of the second Data Partition in order to obtain byte alignment, the next SL packet starting on the next bit of the second Data Partition.
It is another object of the invention to propose a corresponding encoding device.


K.'-
To this end, the invention relates to a device for encoding video data consisting of one or several bitstreams according to the MPEG-4 standard, said device comprising a specific alignment/fragmentation mechanism, according to which, when said video bitstreams are encoded using the syntax mode corresponding to the fragmentation of the Video Object Planes (VOPs) contained in said video data into Video Packets (VPs) and of Video Packets into Data Partitions (DPs), a Video Data Partition is mapped into one or more Sync Layer packets (SL packets), the first Video Data Partition start is always mapped to an SL packet start even if a large Video Data Partition is splitter across several SL packets, and the last SL packet transporting the first Data Partition includes the separation marker and up to 7 subsequent bits of the second Data Partition in order to obtain byte alignment, the next SL packet starting on the next bit of the second Data Partition.
The invention also relates to a coded MPEG-4 signal consisting of at least a , video bitstream obtained at the output of an encoding device in which a specific
alignment/fragmentation mechanism is chosen, according to which, when said video

bitstream(s) is (are) encoded using the syntax mode corresponding to the fragmentation of the Video Object Planes (VOPs) contained m said video data into Video Packets (VPs) and of Video Packets into Data Partitions (DPs), a Video Data Partition is mapped into one or more Sync Layer packets (SL packets), the first Video Data Partition start is always mapped to an SL packet start even if a large Video Data Partition is splitter across several SL packets, and the last SL packet transporting the first Data Partition includes the separation marker and up to 7 subsequent bits of the second Data Partition in order to obtain byte alignment, the next SL packet starting on the next bit of the second Data Partition.
The present invention will now be described, by way of example, with reference to the accompanying drawings in which:
Fig.l gives the main processing layers of a multimedia terminal;
Fig.2 illustrates the alignment/fragmentation mechanism according to the invention.
As said above, the MPEG-4 standard provides standardized ways to represent AVOs of natural or synthetic origin, to compose them together to create compound AVOs that form audio-visual scenes (an audio-visual scene is generally composed of several AVOs,

organized in a hierarchical fashion), to multiplex and synchronize the data associated with AVOs, and to interact with the audio-visual scenes generated or reconstructed at the receiver side. The main processing stages of a multimedia audiovisual tensional allow to render and display such an audiovisual scene is illustrated in Fig.l, where the terminal is a multi-layer structure consisting of the three following layers : a Transmit layer 21, a Flummox layer 22 and an Access Unit layer 23.
The TransMux layer 21 consists of a protection sublayer and a multiplexing sub layer (although it may not be possible to separately identify these sublayers in some TransMux instances, the protection sublayer has a specific interest for providing error protection and error detection tools suitable for the given network or storage medium). This layer 21, not defined in the context of MPEG-4, is in fact an interface to the network or the storage medium and allows to offer transport services matching the requested Quality of Service (QoS). At its output, Lemur streams are available, i.e. a sequence of FlexMux Packets (small data entities consist of a header and a payload).
The FlexMux layer 22, completely specified by MPEG-4, consists of a flexible tool for interleaving data (one or more Elementary Streams into one FlexMux stream) and allows to identify the different channels for the data that have been multiplexed. At the output of said layer 22, SL-Packet zed Streams are available, i.e. a sequence of SL-Packets that encapsulate one elementary stream (SL-Packet, or sync layer Packet = the smallest data entity managed by the next layer 23, or Sync layer, said entity comprising a configurable header and a payload itself consisting of a complete or partial access unit).
The layer 23, or Sync Layer, is provided to adapt elementary streaming data for communication. The elementary streams are conveyed as SL-packet zed streams, and this pocketsize representation additionally provides timing and synchronization information, as well as fragmentation and random access information. This layer 23 is followed by the compression layer 24 which recovers data from its encoded format and allows to perform the necessary operations to decode the encoded signals and reconstruct the original information. This decoded information is then processed (composition, rendering) in view of its presentation (in order to be displayed) or of some user interactions.
According to the invention, illustrated by the alignment/fragmentation mechanism of Fig.2, the following solution is then proposed. When video bitstreams are encoded using the syntax mode corresponding to the fragmentation of VOPs into Video Packets, and of these Video Packets into Data Partitions (DPI, DP2), a Video Data Partition should be mapped into one or more SL packets (SLl, SL2, SL3, SL4, etc) which are byte

aligned (there is a distance of 8 bits or a multiple of 8 bits between the respective first bits of these SL packets, i.e. between the byte boundaries BB of the SL packets). Specifically, it is allowed to split a large Video Data Partition across several SL packets, but the first Video Data Partition DPI start must always be mapped to an SL packet start
Furthermore, since the second Data Partition is useless if the first one is lost but a decoder however needs a marker M to safely identify the end of the first Data Partition DPI, the following alignment rule should be used: the last SL packet transporting the first Data Partition must include the separation marker M (DC marker or Motion Marker, depending on VOP type) and up to 7 subsequent bits of the second Data Partition in order to obtain byte alignment. The next SL packet starts on tribe next bit of the second Data Partition. The specific alignment/firagrnentation thus described allows to map the video elements into the system elements and to avoid in this case any file formation interchange problem or any network problem.


We Claim:
1. A method for encoding video data consisting of one or several bit streams
according to the MPEG-4 standard, characterized in that a specific alignment/fragmentation
mechanism, is chosen, according to which, when said bit streams are encoded using a syntax
mode corresponding to a fragmentation of Video Object Planes (VOPs) contained in said
video data into Video Packets (VPs) and of Video Packets into Data Partitions (DPs), a Video
Data Partition is mapped into one or more Sync Layer packets (SL packets), a first Video
Data Partition start is always mapped to an SL packet start even if a large Video Data
Partition is spitted across several SL packets, and a last SL packet transporting the first Data
Partition includes a separation marker and up to 7 subsequent bits of a second Data Partition
in order to obtain byte alignment, a next SL packet starting on a next bit of the second Data
Partition.
2. A device for encoding video data consisting of one or several bit streams
according to the MPEG-4 standard, said device comprising means (21, 22, 23) configured to

Documents:

in-pct-2002-01765-checorrespondance others.pdf

in-pct-2002-1765-che abstract.jpg

in-pct-2002-1765-che abstract.pdf

in-pct-2002-1765-che claims.pdf

in-pct-2002-1765-che correspondence-others.pdf

in-pct-2002-1765-che correspondence-po.pdf

in-pct-2002-1765-che description(complete).pdf

in-pct-2002-1765-che drawings.pdf

in-pct-2002-1765-che form-1.pdf

in-pct-2002-1765-che form-13.pdf

in-pct-2002-1765-che form-18.pdf

in-pct-2002-1765-che form-26.pdf

in-pct-2002-1765-che form-3.pdf

in-pct-2002-1765-che form-5.pdf

in-pct-2002-1765-che other.pdf

in-pct-2002-1765-che pct search report.pdf

in-pct-2002-1765-che pct.pdf

in-pct-2002-1765-che.jpg


Patent Number 234502
Indian Patent Application Number IN/PCT/2002/1765/CHE
PG Journal Number 29/2009
Publication Date 17-Jul-2009
Grant Date 02-Jun-2009
Date of Filing 25-Oct-2002
Name of Patentee NXP B.V.
Applicant Address HIGH TECH CAMPUS 60, NL-5656 AG EINDHOVEN,
Inventors:
# Inventor's Name Inventor's Address
1 GENTRIC, PHILIPPE PROF. HOLSTLAAN 6 NL-5656 AA EINDHOVEN,
2 NEZOT, BEATRICE PROF. HOLSTLAAN 6 NL-5656 AA EINDHOVEN,
3 RAMANZIN, YVES PROF. HOLSTLAAN 6 NL-5656 AA EINDHOVEN,
PCT International Classification Number H04N7/24
PCT International Application Number PCT/IB2002/00567
PCT International Filing date 2002-02-25
PCT Conventions:
# PCT Application Number Date of Convention Priority Country
1 01400504.5 2001-02-27 EUROPEAN UNION